Ask an owner what a new customer is worth and you usually get the value of the first job. A drain clear is $180, so a new customer is $180, so paying $90 to acquire one feels like giving away half the money.
That framing is wrong in a way that quietly caps growth. The drain clear is the introduction. What you are actually buying is however many jobs follow it, over however many years they stay, plus whoever they refer.
The version of the formula you can actually use
Academic CLV formulas involve discount rates and churn curves. You do not need them. Three numbers you already have get you close enough to make better decisions:
- Average job value. Total revenue divided by number of jobs, over the last year.
- Jobs per customer per year. Total jobs divided by number of distinct customers.
- Average years retained. Harder, but if you have three years of data, look at what share of customers from three years ago are still active.
Multiply them. Average job value × jobs per year × years retained. That is your working lifetime value, and it will almost certainly be several times larger than the number you had in your head.
Why the number changes what you can spend
If a customer is worth $180, then $90 to acquire one is a terrible deal. If a customer is worth $1,400 across four years, $90 is an extraordinary one, and the constraint on your growth is how many you can buy at that price rather than whether you can afford them at all.
This is the practical reason to calculate it. Not because the number is interesting, but because it moves the threshold at which you say yes to a marketing channel. Businesses that price acquisition against the first job systematically underspend, lose bids for attention to competitors who do the arithmetic, and then conclude marketing does not work for their trade.
Where the number differs by trade
| Type of work | Repeat pattern | What drives lifetime value most |
|---|---|---|
| Recurring maintenance | Scheduled, predictable | Retention rate — a single percentage point compounds |
| Seasonal services | Once or twice a year | Whether they remember you next season |
| Emergency and repair | Irregular, need-driven | Being the number they still have saved |
| Large one-off projects | Rarely repeats | Referrals, which may exceed the original job value |
| Mixed repair and install | Small jobs leading to large ones | Whether small jobs are treated as loss leaders or nuisances |
The retention lever is bigger than the acquisition lever
Because lifetime value multiplies rather than adds, extending how long customers stay affects the number more than raising average job value does. A five percent improvement in retention moves lifetime value considerably further than a five percent price rise, and it costs less to achieve.
The mechanics of that are unglamorous: contacting people before they need you rather than after, remembering what happened last time, and not letting a customer go eighteen months without hearing from you. All of which requires the history to be somewhere other than one person's memory.
TARGETS
How to get to a real number this week
- 01Pull last year's revenue and job count
Divide one by the other. That is average job value, and it is usually lower than owners guess.
- 02Count distinct customers, not jobs
Jobs divided by customers gives you frequency. This is the number most businesses have never calculated.
- 03Look three years back
What share of customers active then are active now? That is a usable proxy for retention.
- 04Multiply, then sanity-check
If the answer looks implausibly high, check whether a handful of commercial accounts are distorting it. Calculate residential and commercial separately if so.
Common questions
What is a good customer lifetime value for a service business?
There is no universal figure, because it depends entirely on job value and frequency. The useful comparison is internal: lifetime value against acquisition cost. A ratio of three to one or better, with payback inside a year, generally means you can spend more than you are.
How do I calculate lifetime value without years of data?
Use the data you have and be explicit that it is a floor. If you have eighteen months, calculate value over eighteen months. It will understate the true figure, which is a safer error than overstating it.
Should I calculate this separately for commercial and residential?
Yes, if you do both. They usually have very different job values, frequencies, and retention, and a blended number describes neither accurately.
Does lifetime value include referrals?
It can, and for project-based trades it should, because referrals may be worth more than the original job. Keep it as a separate figure rather than folding it in, so you can see which customers refer.